[XOM-interest] Exposing XMLHandler

Tatu Saloranta cowtowncoder at yahoo.com
Sat Sep 29 01:40:50 EDT 2007


> Being curious... Why do you want to use XOM? I don't use it, but I like
> to look at it. If it doesn't meet your needs, what makes it so
> compelling to push for changes to an obvious <don't-know-what-to-say
> perhaps="fascist | dictator"/>? Why not just dismiss it and use one or
> more of the many other alternatives?

Although I'm not the one asking for support, I think that the reason may be that beyond specific challenges, XOM is a well-designed xml tree model that does many things better than the alternatives. At least that is my opinion, based on observations of 2 out of 3 commonly encountered alternatives.

XOM design has paid careful attention to consistency of the API, which in the end does make things intuitive (via consistency), and specifically does not lead developers to do things they should not.
Some other tree APIs add convenience features that lead users (especially novice ones) into usage patterns that are inappropriate for xml processing (in areas of namespace or white space handling). It's not particularly easy to create a correct and intuitive interface to xml processing, given impedance between logical and common xml and java processing models.

So I think it is probably an indirect compliment that developers try very hard to work around limitations they find, instead of "cutting their losses" so to speak.

Of course individual reasons may be different, this is just my take,

-+ Tatu +-







      ____________________________________________________________________________________
Shape Yahoo! in your own image.  Join our Network Research Panel today!   http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7 




More information about the XOM-interest mailing list